<<

Chapter 4 in , , and southern Stephan Procházka University of Vienna

This chapter covers the Arabic spoken in the region stretching from the Turkish of Mersin in the west to Iraq in the east, including and Syria. The area is characterized by a high degree of linguistic diversity, and for about two and a half millennia Arabic has come into contact with various other Semitic , as well as with Indo-European languages and Turkish. Bilin- gualism, particularly with , Kurdish, and Turkish, has resulted in numer- ous contact-induced changes in all realms of , including and .

1 Current state and historical development

The region discussed in this chapter is linguistically extremely heterogeneous: in it three different Arabic groups, plus several other languages, are spo- ken. The two main Arabic dialect groups are Syrian and Iraqi, the distribution of which does not exactly correspond to the political boundaries of those two . Syrian-type dialects are also spoken in Lebanon, in three of southern Turkey (Mersin, Adana,1 Hatay), and in one village on (cf. Walter, this volume). In Iraq, Arabic is mainly spoken in proper, whereas considerable parts of the mountainous parts of the are Kurdish- speaking. Arabic dialects which are very akin to the Iraqi ones extend into north- eastern Syria and southeastern (for the latter see Akkuş, this volume). These two groups are geographically divided by a third dialect group, which ar- rived in the region with an originally (semi-) nomadic population from northern

1The dialects spoken in Mersin and Adana provinces will henceforth referred to as Cilician Arabic.

Stephan Procházka. 2020. Arabic in Iraq, Syria, and southern Turkey. In Lucas & Stefano Manfredi (eds.), Arabic and contact-induced change, 83–114. Berlin: Science Press. DOI:10.5281/zenodo.3744507 Stephan Procházka

Arabia. Today, this variety preponderates in all villages and most towns between the eastern outskirts of and the right bank of the , and stretching north into the Turkish province of Şanlıurfa. The total number of native Arabic speakers in the whole region is estimated to be 54 million (see Table 1). The dialects of large urban centers like , , Aleppo, and have become supra-regional prestige varieties that are also used in the media and therefore understood by most inhabitants of the respective countries. The situation is very different in Turkey, where the local Arabic is in sharp decline and public life is exclusively dominated by Turkish. Only recently has the position of Arabic in Turkey been socially enhanced by the influx of more than 3.5 million Syrian refugees fleeing the civil war that started in 2011.2 Table 1: Speaker populations for dialects of Arabic

Country Speakers Syria 17,000,000 Lebanon 6,000,000 Iraq 30,000,000 Turkey 1,000,000

Arabic was spoken in the region long before the advent of (Donner 1981: 95), but became the socially dominant language in the wake of the Muslim con- quests in the seventh century CE. From that until the end of the tenth century, when tribes seized large parts of central and northern Syria, there was probably a continuum of sedentary-type dialects that stretched from Mesopotamia to the northeastern Mediterranean (Procházka 2018: 291). During the Mongol sacking of Iraq in 1258, much of the population was killed or ex- pelled. This resulted in far-reaching demographic and linguistic changes as the original sedentary-type dialects were only able to hold ground in Baghdad and the larger settlements to its north. Further south they persisted only among the non-Muslim population. Most of today’s Iraq was re-populated by people who spoke Bedouin-type dialects (mostly coming from the ), which over the centuries have heavily influenced the speech of even most large cities (Holes 2007). Very similar dialects are spoken further south and in the Iranian province of Khuzestan (see Leitner, this volume). The foundation of nation states

2See UNHCR figures at https://data2.unhcr.org/en/situations/syria/location/113.

84 4 Arabic in Iraq, Syria, and southern Turkey after World War One caused a significant decrease in contact between thediffer- ent dialect groups and an almost complete isolation of the Arabic dialects spoken in Turkey.

2 Contact languages

During its two-and-a-half-millennia presence in the region, Arabic has come into contact with many languages, both Semitic and non-Semitic. Those most rele- vant for the topic will be treated in more detail below (for Syria, see also Barbot 1961: 175–177). Akkadian was spoken in southern Iraq until about the turn of the eras, i.e. the first century CE.3 Greek was the language of administration in Greater Syria until the Arab conquest (Magidow 2013: 185–187) and continued to play a role for Orthodox .4 During Crusader , Arabic speak- ers in Syria came into contact with various medieval European languages; and along the Mediterranean coast the so-called (see Nolan, this vol- ume) was an important source for the spread of particularly nautical vocabulary for many centuries (Kahane et al. 1958). Since the nineteenth century, locally re- stricted contacts between Arabic and Armenian and Circassian have existed in parts of Syria and Lebanon.

2.1 Aramaic Aramaic is a Northwest Semitic language and thus structurally very similar to Arabic. Different varieties of Aramaic were the main language in Syria andIraq from the middle of the first millennium BCE and it can be assumed that some contact with Arabic existed even at that time. From the first century CE onwards, the southern fringes of the Fertile Crescent became largely Arabic-dominant and there was significant bilingualism with Aramaic, particularly in the towns along the edge of the steppe, such as , , , and al-Ḥīra (Procházka 2018: 260–262). Though after the Muslim conquests Arabic eventually became thema- jority language, it did not oust Aramaic very quickly: the historical sources sug- gest that Aramaic dominated in the larger towns and the mountainous regions of Syria and Lebanon for a long time. In Iraq, by contrast, the massive influx of into the cities fostered their rapid , while Aramaic continued to be spoken in the countryside (Magidow 2013: 184; 188). But over the centuries,

3For Akkadian lexical influence on Arabic, see Holes (2002) and Krebernik (2008). 4The enormous influence of Modern Greek on the Arabic spoken in the Kormakiti village of Cyprus is discussed by Walter (this volume). For a detailed study, see also Borg (1985).

85 Stephan Procházka the diverse Aramaic dialects became marginalized and, with very few exceptions, were finally relegated to non-Muslim religious minorities, particularly Christians and , in peripheral regions like and the Anti-Lebanon Moun- tains, where Aramaic was prevalent until the eighteenth century (Retsö 2011). Western Aramaic is still spoken in three Syrian villages, the best known of which is .5 There also remain speakers of Neo-Aramaic in northern Iraq.6 It is hard to establish the degree of bilingualism in the past, but it can be as- sumed that it was mostly Aramaic L1 speakers who had a command of Arabic and not vice versa. In the present time, nearly all remaining Aramaic speakers in Syria are fluent in Arabic. In Iraq this is mainly true of those living in theplain just north of (Arnold & Behnstedt 1993; Coghill 2012: 86). The influence of different strata of Aramaic on spoken Arabic is a long debated issue, various scholars rating it from considerable to negligible (Hopkins 1995: 39; Lentin 2018: 199–204).

2.2 Persian and Kurdish For many centuries, Arabic and the two Western Persian and Kurdish have influenced each other on different levels. Persian-speaking com- munities existed in medieval Iraq, and economic and cultural contacts between Mesopotamia and have continued to the present (cf. Gazsi 2011). An impor- tant factor of are the holy shrines of the in Kerbela, , and other Iraqi cities, which have always attracted tens of thousands of Persian- speaking Shiites every year. Intensive contacts between speakers of Kurdish and Arabic have existed since at least the tenth century, particularly in Northern Iraq, northeast Syria, and southeast Anatolia (see Akkuş, this volume). Until their ex- odus in the early 1950s, the Arabic-speaking Jewish communities which existed in usually had a native-like command of Kurdish (Jastrow 1990: 12). Due to the multilingual character of the region, bilingualism in Kurdish and Arabic is still relatively widespread, particularly in urban settings, though with usually much more fluent in Arabic than the other way around.7 However, for obvious reasons, little linguistic research has been done in Iraq for decades, which makes it impossible to give up-to-date information about the linguistic situation in ethnically-mixed cities like .

5The village heavily suffered from the jihadist occupation of 2013–2014, but after government troops had retaken control over the region, many inhabitants returned and began its recon- struction (cf. the reports collected at http://friendsofmaaloula.de/). 6See Coghill (2012) and http://glottolog.org/resource/languoid/id/nort3241. 7With significant exceptions in some parts of southeast Anatolia; see Akkuş (this volume).

86 4 Arabic in Iraq, Syria, and southern Turkey

2.3 Ottoman and Modern Turkish Contacts between spoken Arabic varieties and various existed from the ninth century onwards. These early contacts, however, left hardly any traces in Arabic except for a handful of . In the sixteenth century, the Ottomans established their rule over most Arab lands, including Syria, Lebanon, and Iraq. This domination lasted four hundred years, until World War One. Par- ticularly in the provinces of Aleppo and Mosul, there was a relatively high per- centage of Turkish speakers and probably a significant degree of bilingualism.8 As the language of the ruling elite, Turkish had high prestige and therefore was at least rudimentarily spoken by many inhabitants of those regions, especially urban . The collapse of the put an abrupt end to Turkish– Arabic contacts, which today remain intensive only among the Arabic varieties spoken within the borders of Turkey itself, where most Arabic speakers are flu- ent in Turkish, the dominant language in all contact settings. In some areas of Syria and in northern Iraq, the Arabic-speaking population by side with several hundred thousand speakers of Turkish and Azeri Turkish, who call themselves . Unfortunately, no reliable data on the sociolinguistic settings and the degree of bilingualism exist for those areas. Again, it can be assumed that most of the Turkmens in both countries are dominant in Turkish, but use Arabic as a .

2.4 French and English After World War One, Syria and Lebanon stayed under the French mandate and Iraq under the British mandate until they reached independence.9 French is still widely spoken as a second language in Lebanon, especially by Christians. In Iraq, English has maintained its position as by far the most important – a fact which was reinforced by the US military occupation from 2003 to 2010.

2.5 Intra-Arabic contacts Contacts between different Arabic varieties, for instance between speakers ofru- ral and urban dialects, happen on an everyday basis and often trigger short-term accommodation without leading to long-lasting changes. The situation is differ- ent with regard to the enduring contacts between the Bedouin and the sedentary

8See Wilkins (2010: xv) for Aleppo. Koury (1987: 103) maintains that Aleppo’s hinterland was culturally even more Turkish than Arab. For Mosul, see Shields (2004: 54–55). 9Iraq in 1932, Lebanon in 1943, Syria in 1946.

87 Stephan Procházka populations, whose dialects differ from each other considerably.10 Such contacts are most intense at the periphery of the Syrian steppe and along the middle Eu- phrates, where scattered towns with sedentary dialects like Palmyra, Deir ez-Zor and Hit are surrounded by an originally nomadic population. Though the no- madic way of life has been abandoned by most of them, they still speak Bedouin- type Arabic dialects. As the nomads were, for many centuries, both socially and economically dominant, speakers of sedentary dialects often adopted linguistic features from more prestigious Bedouin (though reverse instances are also found; cf. Behnstedt 1994a: 421). Due to the historical circumstances mentioned in §1, also had a strong linguistic impact on Iraqi dialects. In Baghdad the sedentary dialect of the Muslim population has been gradually Bedouinized due to massive migration from the countryside to the city (Palva 2009). The Chris- tian and, in former times, Jewish inhabitants, on the other hand, preserved their original sedentary-type dialects because they had much less contact with the Muslim newcomers.

3 Contact-induced changes

Change induced by contact with Aramaic almost exclusively happened through imposition, that is, by Aramaic speakers who had learned Arabic as a second lan- guage and later often completely shifted to Arabic. This explains the relatively numerous phonological changes and pattern replications in syntax. Lexical trans- fers from Aramaic certainly were also made by Arabic-dominant speakers, par- ticularly in semantic fields like agriculture that included concepts forthe mostly animal-breeding Arabs. The same is true for transfers from Greek, for which a very low level of bilin- gualism can be assumed. Thus we find only matter replication (in the sense of Sakel 2007) in the form of loanwords, mostly in domains where lexical gaps in older layers of spoken Arabic are likely. In the case of transfer from Kurdish, bilingualism is much more widespread among speakers of the source language, suggesting imposition. This might ex- plain some of the phonological changes discussed in §3.1.2, as speakers domi- nant in the source language tend to preserve its phonological features (Lucas 2015: 532). The relatively small number of instances of lexical matter replication is probably the result of the fact that Arabic has long been regarded as the more prestigious by speakers of both the source and the recipient language.

10Since these two speech communities differ from each other in so many ways, it is a relatively robust approach to rate the following features as results of dialect contact and not mere vari- ation (cf. Lucas 2015: 533).

88 4 Arabic in Iraq, Syria, and southern Turkey

The numerous loanwords from Persian into Iraqi Arabic may well be the re- sult of matter replication by agents who were dominant in the recipient language Arabic. Starting with the rule of the Abbasid caliphs in the eighth century CE and continuing to the present, Iranian material culture and cuisine often had a great impact on neighbouring Mesopotamia. There were also many intellectu- als, among them praised of Arabic prose, who were actually Iranians and hence knew both languages. Frequent contacts on the everyday level caused ad- ditional borrowing of ordinary vocabulary and the retention of sounds that are replaced in Persian loans found in or other dialects.11 Changes induced by contact with may have happened most- ly through Arabic-dominant speakers. The current situation of Arabic speakers in Turkey is, however, very different, because at least the last two generations have acquired Turkish as an L2 or even as a second L1 at very young age. Thus, at least some of the contact phenomena described in the following paragraphs may be examples of linguistic convergence (see Lucas 2015: 525). French and English have largely remained typical “foreign languages” learned at school or in business with a considerable amount of bilingualism only in some urban settings of Lebanon, particularly Beirut. The agents of change are certainly dominant in the recipient language. The distinction between the two transfer types is not always clearly discern- ible in case of intra-Arabic contact-induced changes. In the towns of the Syrian steppe and the middle the agents of change were mostly the seden- tary population who adapted their speech towards the norms of the socially more prestigious Bedouin. However, there has always been inter-marriage, and Bedouins often settled in towns and may well have adopted features from the local sedentary variety. Especially in cases like Muslim Baghdadi (see §1), we may assume with good reason that the Bedouin character of today’s variety de- veloped through both imposition and borrowing.

3.1 3.1.1 Aramaic-induced changes It has been hypothesized that several phonological features of the Syrian and Lebanese dialects are due to the contact-induced influence of Aramaic. But in the case of the shift from interdental to postdental (/ð/ >/d/; /θ/ > /t/; /ð̣/ > //) this is unlikely because: (i) this sound change is common crosslinguistically; (ii) it does not occur in all dialects of the region; and (iii) it is found in many other Arabic dialects without an Aramaic substrate.

11The phonological changes are not, however, only the result of Persian influence (cf.3.1.2 § ).

89 Stephan Procházka

A phonotactic characteristic of most dialects spoken along the Mediterranean, from in the north to Beirut in the south, is that all unstressed short (including /a/) in open are elided,12 whereas in other dialects east of only /i/ and /u/ in this position are consistently dropped.

(1) Cilician Arabic (Procházka 2002a: 31–32; 130) (OA) *raṣāṣ > rṣāṣ ‘lead, plumb’ OA *miknasa > mikinsi ‘broom’13 *fataḥ-t > ftaḥt ‘I opened’

Because this rule corresponds to the phonotactics of Aramaic and is otherwise not found to the same degree except in Maghrebi dialects (cf. Benkato, this vol- ume), pattern replication is likely, though cannot be proved.14 In roughly the same region, except Cilicia and many dialects of Hatay,15 the /ay/ and /aw/ are only preserved in open syllables, but monophthong- ized to /ē/ and /ō/ respectively in closed syllables. In some regions, for instance on the island of , both diphthongs merge to /ā/ in closed syllables (Behnstedt 1997: map 31).

(2) Arwad, western Syria (Procházka 2013: 278) OA *bayt, *baytayn > bāt, baytān ‘house, two houses’ OA *yawm, *yawmayn > yām, yawmān ‘day, two days’ OA *bayn al-iθnayn > bān it-tnān ‘between the two’

Likewise, in older layers of Aramaic, diphthongs were usually monophthong- ized in closed syllables (for Syriac see Nöldeke 1904: 34), which makes imposition by L1 speakers of Aramaic rather likely (Fleisch 1974a: 227). Another striking phenomenon is the split of historical /ā/ into /ō/ and /ē/ that is found in scattered areas of the , particularly northern Lebanon, around the Syrian port of Tartous, the Qalamūn Mountains, and the exclusively Christian town of Maḥarde on the .16 Because in many varieties of Aramaic the old Semitic /ā/ is reflected as /ō/, it could be assumed that Aramaic speak- ers transferred their peculiar pronunciation to Arabic when learning it. Fleisch

12Therefore, Cantineau (1960: 108) called them parlers non différentiels – a term still very often applied in Arabic dialectology – as they make no distinction in the treatment of the three short vowels. 13With insertion of an epenthetic /i/ to avoid a sequence of three . 14Cf. Diem (1979: 47); Arnold & Behnstedt (1993: 69–71); Weninger (2011: 748). 15Where this phenomenon occurs only in Alawi villages (Arnold 1998: 84). 16For details cf. Behnstedt (1997: map 32). The conditioned shift /ā/ > /ō/ is also found inand around in Turkey (Procházka 2002a: 37–38).

90 4 Arabic in Iraq, Syria, and southern Turkey

(1974b: 49) rejected the hypothesis of an Aramaic influence, arguing that the con- ditioned distribution of the two is merely a further development of the [ɒ] : [æ] split widely attested for Lebanon and parts of western Syria. However, in the Syrian Qalamūn Mountains there are dialects with an unconditioned shift (Behnstedt 1992), and this is precisely the region where the shift from Aramaic to Arabic occurred relatively late, probably after a long phase of bilingualism. Inthe town of Nabk, for instance, one can infer that the former Aramaic speaking in- habitants would have simply turned every /ā/ into /ō/ – except those which long before had become [ē] (or [ɛ̄]) as a result of the so-called conditioned imāla (i.e. the tendency of long /ā/ to be raised towards [ē] or even [ī] if the word contains an /i/ or /ī/).17 Example (3) clearly shows that the distribution of the allophones is not conditioned by the consonantal environment.

(3) Nabk, Syria (Gralla 2006: 20) OA *ṭābiḫ > ṭɛ̄beḫ ‘cooking’ vs. OA *ṭālib > ṭōleb ‘student’ OA *ḥāmil > ḥɛ̄mel ‘pregnant’ vs. OA *ḥāmiḍ > ḥōmeḍ’ ‘sour’

In these cases Aramaic influence seems plausible. For the region of it may be assumed that Aramaic bilinguals from the adjacent mountains used [ō] instead of [ā] when speaking Arabic and thus reinforced the already existing [ɒ] : [æ] split.18

3.1.2 The “new” /č/, //, and /p/ Consonantal phonemes that are originally alien to Arabic are found in all Ara- bic dialects spoken in Turkey (see also Akkuş, this volume), northern Syria, and Iraq. These are the unvoiced /č/, the voiced19 /g/, and the unvoiced /p/, the latter mainly used in Iraq. The emergence of these sounds was very likely contact-induced, but it is often impossible to discern which language triggered each development: all three sounds are found in Persian, Kurdish, Turkish, and the Lingua Franca. For the dialects of Cilicia, Hatay and Syria, the main source language doubtless was Turkish. The sound /p/ in the Iraqi dialects was probably first introduced through contact with Persian and Kurdish, and then reinforced

17Cf. Arnold & Behnstedt (1993: 68). 18For discussion see Fleisch (1974b: 48–50; 1974a: 133–136), Diem (1979: 45–46); Behnstedt (1992); Arnold & Behnstedt (1993: 67–68); Weninger (2011: 748). 19The sound [g] is prevalent in whole Syria and Lebanon but seems to have phonemic status only in the north (Sabuni 1980: 26). For further examples and discussion see Ferguson (1969). This “foreign” /g/ must therefore be differentiated from the /g/ which is the regular reflexof OA *q. The latter development is found in many Bedouin-type dialects.

91 Stephan Procházka by Ottoman Turkish. In the Bedouin-type dialects of the region, the phonemes /č/ and /g/ are not products of contact-induced change but occur due to internal sound changes, unvoiced /č/ as a conditioned affricated variant of /k/ and /g/ as the ordinary reflex of OA *q. Thus, it can be assumed that over the centuries speakers of the sedentary - lects of Iraq and Syria borrowed either from other languages or from varieties words that possess these two sounds, which subsequently were fully incorporated into the phonemic inventory. This development may have been facilitated by the fact that the three sounds /č/, /p/, and /g/ are not funda- mentally unfamiliar to Arabic, but are the voiceless/voiced counterparts of the well-established phonemes /ǧ/, /b/, and /k/. It seems no accident that the new sound /č/ is much more often found in dialects that have preserved the affricate /ǧ/ than in those where it has shifted to /ž/, as illustrated in examples4 ( ) and (5).

(4) Aleppo (Sabuni 1980: 205–210) čanṭāye ‘handbag’ (Turkish çanta) čwāl ‘sack’ (Turkish çuval) čāy ‘tea’ (Turkish çay) gaǧaleg ‘nightgown’ (Turkish gecelik)

The words given in (4) are usually pronounced with [š] instead of [č] in the central Syrian and Lebanese dialects where contact with Turkish was less intense and /ǧ/ is reflected as /ž/.20

(5) Mosul (own data) ṣūč ‘fault’ (Turkish suç) pāča ‘stew of sheep and cow legs and innards’ (Kurdish/Persian pāče) zangīn ‘rich’ (Turkish zengin)

Once integrated into the phonological system, these sounds not only enabled easier integration of loanwords from other languages like French and English (see §2.4), but sometimes also resulted in the spread of assimilation-induced allo- phones from single words to the whole paradigm or even . In Aleppo one finds *yəkdeb > yəgdeb ‘ lies’, due to assimilation. The g subsequently was transferred to other words derived from the root: gadab ‘he lied’, gədbe ‘lie’, and gaddāb ‘liar’ (Sabuni 1980: 26, 209).

20Cf. Behnstedt (1997: maps 18, 19, 25). For details and more examples see Sabuni (1980: 205–210), who lists all words with č/g in Aleppo, and Procházka (2002b: 185) for Cilician Arabic.

92 4 Arabic in Iraq, Syria, and southern Turkey

Speakers of sedentary dialects who had everyday contact with Bedouins – for example the inhabitants of Deir ez-Zor and Khatuniyya – first integrated /č/ and /g/ into their phonemic inventory through the borrowing of typically Bedouin vocabulary such as dabča ‘a Bedouin dance’ (Khawetna; Talay 1999: 29) and ṭabga ‘milk-bowl’ (Soukhne; Behnstedt 1994b: 310). These sounds then entered other fields of the lexicon, which led to unpredictable distribution, including doublets, as in (6)–(8).

(6) Khawetna (Talay 1999: 28–31) gəṣṣa ‘forehead’, but qəṣṣa ‘story’ (OA *quṣṣa / *qiṣṣa) dīč ‘rooster’ (OA *dīk) (7) Deir iz-Zor (Jastrow 1978: 42–43). gāʕ ‘soil’ (OA *qāʕ) čam ‘how much?’ (OA *kam) (8) Baghdad (Palva 2009: 18–19) guffa ‘large basket’ (OA *quffa), but quful ‘lock’ (OA *qufl) ʕigab ‘to pass’, but ʕiqab ‘to follow’ (both OA *ʕaqab)

The opposition /k/ : /č/ has even entered morphology, particularly with the 2sg : ʔabū-k ‘your (sg.m) father’ vs. ʔabū-č ‘your (sg.f) father’. In the Syrian oasis of Soukhne, long-term contact with speakers of Bedouin dialects caused a chain of phonetic changes: first /k/ shifted to /č/, which originally was thereflex of OA /ǧ/; then /č/ (< /ǧ/) shifted further to /ts/, which has become a unique feature of the local dialect. The unconditioned shift from /k/ > /č/, which isnot found in the Bedouin dialects, in turn caused a shift from /q/ > /k/.21

(9) Soukhne (Behnstedt 1994b: 226, 344, 357, 360) kirbi ‘water-skin’ (< OA *qirba, Bedouin girba) čalb ‘dog’ (< OA *kalb, Bedouin čalib) čurr ‘donkey foal’ (< OA *kurr, Bedouin kuṛṛ) tsubn ‘cheese’ (< OA *ǧubn, Bedouin ǧubun)

3.2 Morphology 3.2.1 Diminutive The Aramaic diminutive -ūn has become restrictedly productive in Iraqi Arabic (Masliyah 1997: 72), as illustrated in (10). In Syria and Lebanon it is only

21See Behnstedt (1994b: 4–11) for details.

93 Stephan Procházka found in fossilized forms such as šalfūn ‘young cockerel’ and qafṣūne ‘little cage’. Such kinds of morphological transfer are usually triggered by lexical borrowing. Thus, it may be assumed that this suffix spread from loanwords like šalfūne ‘small knife blade’ < Aramaic šelpūnā ‘little knife’ (cf. Féghali 1918: 82).22 (10) Iraq (Masliyah 1997: 72) darb ‘road’ > darbūna ‘’ gṣayyir ‘short’ < gṣayyrūn ‘very short’ mḥammdūn hypocoristic form of the name Muḥammad

3.2.2 Morphological templates Syrian and Lebanese dialects exhibit a few word patterns (templates) that are attested for OA (and other dialects) but seem to have become widespread through contact with Aramaic due to their frequency in the latter. These are the verbal pattern šaC1C2aC3 and the (primarily diminutive) nominal patterns C1aC2C2ūC3 23 and C1aC2C3ūC4. An example of the first is šanfaḫ ‘to puff up’, related to nafaḫ ‘to blow up’ (Féghali 1918: 83; cf. Lentin 2018: 201 for further discussion); the nominal forms are illustrated in (11) and (12). (11) Aleppo (Barthélemy 1935: 104, 158, 851) ǧaḥḥūš ‘little donkey’ (related to ǧaḥš ‘young donkey’) ḥassūn ‘goldfinch’ (related to the personal name ḥasan) namnūme ‘small louse’ (cf. naml ‘ants’)

The pattern C1aC2C2ūC3(i) is still productive in the whole region, including the Bedouin dialects, to derive hypocoristic forms from personal names: (12) fāṭma > faṭṭūma ḥalīme > ḥallūma aḥmad/mḥammad > ḥammūdi

3.2.3 In all Syrian and Lebanese dialects, as well as in Anatolia, the 2pl and 3pl pro- nouns exhibit an /n/ in place of the /m/ that is found in other Arabic dialects, which makes them look as if they were reflexes of OA feminine forms (Table 2).

22This must be a very old borrowing because the suffix is also found in the Gulf dialects (e.g. ḥabbūna ‘a little’ Holes 2002: 279) and even in (Singer 1984: 496), where direct Aramaic influence can be excluded. 23For the latter two see Corriente (1969) and Procházka (2004).

94 4 Arabic in Iraq, Syria, and southern Turkey

Table 2: 2pl and 3pl pronouns

Damascus OA pl.f Syriac pl.m 2pl ʔəntu / -kon ʔintu/-kom ʔantunna / -kunna ʔatton / -kon 3pl hənne(n) / -hon humme / -hom hunna / -hunna hennon / -hon

Because generalization of the feminine is unlikely,24 these forms have often been explained as a contact-induced change. In Aramaic the corresponding pro- nouns also have /n/ (for Syriac see Muraoka 2005: 18). In particular, the 3rd per- forms with final -n exactly mirror the Aramaic pattern, but lack a plausible intra-Arabic . Thus imposition seems plausible. Nevertheless, substra- tum influence has been doubted, particularly because of the infrequent evidence of n-pronouns in other regions.25

3.2.4 Vocative suffixes The suffixes -o (in the west of the region) and -u (in the east) can be attached to various kinship terms and given names when used for direct address, usually hypocoristically.26

(13) (own data) šnōnak ḫayy-o? ‘Brother, how are you?’ ǧidd-o ‘Grandfather!’ ʕamm-o ‘(paternal) Uncle!’ ḫāl-o ‘(maternal) Uncle!’

In Syria the suffix is also added to female nouns: ʕamm-t-o ‘(paternal) Aunt!’ and ḫāl-t-o ‘(maternal) Aunt!’, whereas in Iraq the corresponding forms end in -a: ʕamm-a, ḫāl-a. Since this suffix has no overt Arabic etymology, it has been assumed tobea borrowing of the Kurdish vocative -o (e.g. Grigore 2007: 203). The Persian suf- fix -u also forms affective diminutives,27 which would make Persian influence

24This is mainly because the feminine forms are only used for addressing groups of females, whereas the masculine forms may also refer to a mixed group. Therefore, the masculine forms are certainly more frequent. In all Arabic dialects except those mentioned above, the gender- neutral plural forms are clearly derived from the historical masculine. 25See Owens (2006: 244–245) and Procházka (2018: 283–284) for details. 26See also Ferguson (1997: 187). 27E.g. pesar-u ‘kid’; ʕamm-u is even the common word for ‘uncle’ (Perry 2007: 1011).

95 Stephan Procházka possible, at least for Iraq.28 However, the distribution of this feature extends far beyond even indirect contact with Kurdish or Persian,29 though reinforcement and influence on the phonology may be possible for certain regions. Similar end- ings in Aramaic (Fassberg 2010: 88–89) and Ethiopian (Brockelmann 1928: 122) suggest a common Semitic origin (see also Pat- 2017: 463–465).

3.2.5 Turkish derivational suffixes All dialects of the region have incorporated the Turkish suffix -ci [ʤi] into their nominal morphology, as illustrated in (14) and (15). This suffix has become pro- ductive and is therefore a good example of morphological matter borrowing (Gar- dani et al. 2015). It is widely used for expressing professions, occupations, and habitual actions – the latter overwhelmingly pejorative, or at least humorous. In Iraqi dialects the suffix is reflected as -či, which corresponds to its pronuncia- tion in the regional Turkish varieties. In the other varieties, it follows the usual development of *ǧ, which means that it is realized as -ǧi or -ži. (14) Syria/Damascus (own data) kahrab-ži ‘electrician’ (kahraba ‘electricity’) nəswān-ži ‘womanizer’ (nəswān ‘women’) maškal-ži ‘troublemaker’ (məšəkle ‘problem’) (15) Iraq (Procházka-Eisl 2018: 40–44) pančar-či ‘tire repairman’ (pančar ‘puncture’) mharrib-či ‘ trafficker’ (mharrib ‘one who helps s.o. to escape’) ʕarag-či ‘drunkard’ (ʕarag ‘aniseed brandy’) The suffix clearly fills a morphological gap, because it enables morphologi- cally transparent derivation even from loanwords, by preserving the basic, im- mediately recognizable word – in contrast to the Arabic C1aC2C2āC3 pattern or , which are derived from the root (for details see Procházka-Eisl 2018). To a lesser extent other Turkish suffixes have enhanced the morphological devices of the dialects treated here,30 specifically the relative suffix -li, the priva- tive suffix -siz, and the abstract suffix -lik, which is reflected as -loɣiyya in Iraq,

28In the Iraqi dialects the is -u, e.g. ʕamm-u, ḫāl-u and ǧidd-u (Abu-Haidar 1999: 145). 29The suffix is, for instance, attached to given names for endearment in the Gulf dialects, cf. Holes (2016: 128). The address forms ya ʕamm-u, ya ḫāl-u ‘uncle’, gidd-u ‘grandfather’, sitt-u ‘grandmother’ are used in , where hypocoristic variants of given names are likewise at- tested, e.g. mīšu for hišām (Woidich 2006: 109). The suffix -o/-u in address forms is also attested in eastern (Stefano Manfredi, personal communication), and in the ; Prunet & Idrissi (2014: 184) provide a list of such nouns for . 30See Halasi-Kun (1969: 68–71); Sabuni (1980: 168); Masliyah (1996); Procházka (2002a: 186).

96 4 Arabic in Iraq, Syria, and southern Turkey i.e. with the Arabic abstract morpheme affixed. For the most part these suffixes appear in Turkish loanwords, e.g. Cilicia ṣiḥḥat-li (< Turkish sıhhatlı) ‘healthy’, raḥaṭ-ṣīz (< Turkish rahatsız) ‘uncomfortable’. Only in Iraq have they gained a certain degree of productivity, particularly -sizz and -loɣiyya:

(16) Iraq (Masliyah 1996: 293–294) muḫḫ-sizz ‘stupid, brainless’ ḥaya-sizz ‘shameless’ ḥaywān-loɣiyya ‘ignorance’ (lit. ‘animal-ness’) zmāl-loɣiyya ‘stupidity’ (lit. ‘donkey-ness’)

3.2.6 Light- constructions Arabic dialects spoken in Turkey not infrequently use light-verb constructions (in Turkish grammar mostly called phrasal ) which consist of the verb ‘to do’ plus a following noun (see also Akkuş, this volume). Such compound verbs are very frequent in Turkish (and Kurdish) and enable easy integration of foreign vocabulary into the verbal system. The light verbs found in the Arabic dialects show that this formation is a case of selected pattern replication because, first, not all examples are exact copies of the Turkish model, and second, the follows the Arabic VO rather than the Turkish OV pattern:

(17) –Urfa (own data) sāwa qaza (Turkish yapmak) ‘to have an accident’ sāwa ʕēš (in Turkish not a phrasal verb, but pişirmek) ‘to cook’ (18) Cilician Arabic (Procházka 2002a: 198) sawwa zarar (Turkish zarar vermek) ‘to harm’ sawwa ḫayir (Turkish hayır işlemek) ‘to do a good deed’

3.2.7 Intra-Arabic dialect contact Concerning intra-Arabic contact, here we see that this has led to the adoption of typical Bedouin-type pronouns into sedentary dialects (cf. Palva 2009: 27–29), e.g.:

(19) Baghdad, Deir ez-Zor, Soukhne ʔəḥna for nəḥna 1pl (20) Baghdad ʔāni for ʔana 1sg

97 Stephan Procházka

In addition, as shown in Table 3, virtually all the eastern sedentary dialects of Syria have copied the typical Bedouin-type active participles of the verbs ‘to eat’ and ‘to take’, which exhibit initial m- (Behnstedt 1997: map 175).

Table 3: Active participles of the verbs ‘to eat’ / ‘to take’

Bedouin Soukhne Palmyra Damascus māčil / māḫið mīčil / mīḫið mākil / māḫið ʔākel / ʔāḫed

Finally, in a few places intensive mutual contact has resulted in an interdialect (Trudgill 1986: 62) with completely new forms, such as the 3pl.m inflectional suffix -a in the Syrian village of Ṣōrān (Behnstedt 1994a: 423–425), as shown in Table 4. Table 4: 3pl.m inflectional suffixes – ‘they said’

Bedouin Sedentary Ṣōrān gāḷ-am qāl-o qāl-a

3.3 Syntax 3.3.1 Changes due to contact with Aramaic 3.3.1.1 doubling In all but the Bedouin-type dialects of the region, two constructions exist which both use an anticipatory and the preposition l- ‘to’: (i) a construction involving analytical marking of a definite direct , as in(21–23); and (ii) a construction involving analytic attribution of a noun, as in (24). The frequency and constraints of these two cases of clitic doubling show great variety, but in general the usage of construction (i) is restricted to specific objects, particularly elements denoting human , and construction (ii) is mostly found with in- alienable possession, particularly kinship. A detailed discussion of both features is found in Souag (2017).

98 4 Arabic in Iraq, Syria, and southern Turkey

(21) Damascus (Berlinches 2016: 144) ḥabbēt-o la-ʕamər .prf.1sg-3sg.m to-Amr ‘I loved Amr.’ (22) Baghdad, Christian (Abu-Haidar 1991: 116) qaɣētū-nu l-əl-əktēb read.prf.1sg-3sg.m to-def- ‘I read the book.’ (23) Cilician Arabic (ʕalā instead of l-; Procházka 2002a: 158) biyḥibb-u ʕala ḫāl-u love.impf.ind.3sg.m-3sg.m on uncle-3sg.m ‘He his (maternal) uncle.’ (24) Baghdad, Christian (Abu-Haidar 1991: 116) maɣt-u l-aḫū-yi wife-3sg.m to-brother-obl.1sg ‘my brother’s wife’ Though the preposition l- is sometimes attested in Classical Arabic for intro- ducing direct objects and is common even in for analytic noun annexation, there are good arguments that the two constructions are pat- tern replications of an Aramaic model.31 For one thing, they do not have direct parallels either in OA or in dialects which lacked contact with Aramaic. Example (25) shows that both constructions have striking parallels in especially the later eastern varieties of Aramaic (Rubin 2005: 94–104). (25) a. Syriac (Rubin 2005: 100) bnā-y l-bayt-ā build.prf.3sg.m-3sg.m to-house-def ‘He built the house.’ b. Syriac (Hopkins 1997: 29)32 šm-ēh l-gabr-ā name-3sg.m to-man-def ‘the name of the man’ 31Not discussed here are two variants of construction (i), one without the suffix and the other without the preposition (cf. Lentin 2018: 203). Among the many studies that are in favor of Ar- amaic influence are Contini (1999: 105); Blanc (1964: 130); and Weninger (2011: 750). Diem (1979: 47–49) and Lentin (2018) are more skeptical. Souag (2017: 52) suggests that at least “the initial stages of the development of clitic doubling in the Levant derive from Aramaic substratum influence, but the current situation also reflects subsequent Arabic-internal developments”. 32The same pattern using the linker d- is more common.

99 Stephan Procházka

3.3.1.2 fī ‘can’ In the entire western part of the region including southern Turkey, the preposi- tion fī ‘in’, together with a pronominal suffix, is used to express a capability, as in (26). This has a striking parallel in the modern Aramaic ʔīθ b- ‘there is in’ ~ ‘be able’ (Borg 2004: 52).

(26) Damascus (Cowell 1964: 415) fī-ni sāʕd-ak əb-kamm lēra in-1sg help.impf.1sg-2sg.m with-some pound ‘Can I help you with a few pounds?’

3.3.1.3 Specific indefinite šī A final example of possible Aramaic influence is the Syrian particle šī that mainly indicates partial specifity, as in (27). It might be a pattern replication of the West- ern Neo-Aramaic form mett, used with the same function (Diem 1979: 49). What reduces the likelihood of imposition by Aramaic speakers is the existence of a in which is used with almost the same function.33

(27) Damascus (own data) hnīk fī šī ʕamūd there exs indf column ‘There is some column.’

3.3.2 Changes due to contact with other languages 3.3.2.1 Indefiniteness A hallmark of both sedentary and Bedouin-type Iraqi dialects is that reflexes of the noun ‘individual (thing or person)’ are used to mark different kinds of in- (Blanc 1964: 118–119). The same form with the same indefinite function is found in in the Iranian province of Khuzestan, and in all Arabic speak- ing language islands of , i.e. Khorasan, , and , as illustrated in (28).

(28) Kirkuk (own data) taʕrif-lak fadd ṭabīb bāṭiniyye know.impf.2sg.m-dat.2sg.m indf doctor internal ‘Do you know a doctor of internal ?’

33Cf. Brustad (2000: 19, 26–27) and Wilmsen (2014: 51–53).

100 4 Arabic in Iraq, Syria, and southern Turkey

It is very likely that the noun fard has developed into a kind of indefinite arti- cle under the influence of other areal languages, particularly Turkish, Turkmen, Persian, and Neo-Aramaic. However, in contrast to all contact languages, Iraqi Arabic has not grammaticalized the numeral ‘one’ (wāḥəd), but fard. This clearly indicates that this feature is a case of pattern replication. There are many paral- lels in the functions of the indefinite articles (such as marking pragmatic salience, semantic individualization, approximation with numerals). Moreover, in all lan- guages they are not fully systematized as a grammatical category as their usage is often optional. In the dialects of the Jews of Kurdistan the definite article is often omitted in subject position – a flagrant imitation of the Kurdish model (see also Akkuş, this volume, for some Anatolian dialects).

(29) Kurdistan Arabic (Jastrow 1990: 71) baʕdēn mudīra baʕatət ḫalf-na then director send.prf.3sg.f after-1pl ‘Then the director sent for us.’

3.3.2.2 m-bōr ‘because, in order to’ An interesting case of calquing which shows the difficulty of distinguishing be- tween borrowing and imposition (see Manfredi, this volume) is the m-bōr ‘because, in order to’. It exhibits both matter and pattern transfer, as it is a copy of Kurdish ji ber (ku). In the actual form the Kurdish ji ‘from’ was replaced by the Arabic equivalent m-(Jastrow 1979: 64).

3.3.2.3 Evidentiality Syntactic change because of contact with Turkish is restricted to the Arabic dia- lects spoken in Turkey. In Cilicia and the Harran–Urfa region, active participles express evidentiality, that is, they are used in utterances where a speaker refers to second-hand information. As evidentiality is not a common category in Semitic, it is very likely that the bilingual Arabic speakers of those regions copied this lin- guistic category from Turkish. In Turkish, any second-hand information is obliga- torily marked by the verbal suffix -mış, whose second function besides evidential- ity is to express stativity and perfectivity. The latter two functions are assumed by the active in many Arabic dialects, including those in here. Thus, we can suppose that the stative/perfective function, which is shared by both Arabic active participles and the Turkish suffix -mış, was likely the starting

101 Stephan Procházka point of the development that led to the additional evidential function of Arabic participles. The fact that evidentials seem to spread readily through language con- tact (Aikhenvald 2004: 10) makes Turkish influence even more probable.34 The example in (30) illustrates how the speaker uses perfect forms for those parts of the narrative he witnessed himself, and participles for secondhand information (perfect forms italic, participles in bold face).

(30) Harran–Urfa (Procházka & Batan 2016: 465) ʔiḥne b-zimānāt čān ʕid-na ǧār b-al-maḥalle huwwa māt ərtiḥam əngūl-lu šēḫ mǝṭar […] nahāṛ rabīʕ-u wāḥad ʕāzm-u ʕala stanbūl rāyiḥ maʕzūm ʕala stanbul māḫið šēḫ mǝṭar əb-sāgt-u ‘Once we had a neighbor in our quarter. He died; he passed away. We called him Mǝṭar. One day somebody invited his friend to . As he was invited he went to Istanbul and he took Sheikh Mǝṭar with him.’

3.3.2.4 Comparative and superlative In most Arabic dialects that are spoken in Turkey, comparatives and superlatives may be expressed by means of the Turkish particles daha and en, respectively, followed by the simplex instead of the elative form of the (cf. Akkuş, this volume). As for comparatives, the use of such constructions is rather restricted, while, at least in Cilician Arabic, they are relatively frequent for the superlative.

(31) Harran–Urfa (own data) daha zēn ṣārat more good become.prf.3sg.f ‘It has become better.’ (32) Cilician Arabic (Procházka 2002a: 155) mīn en zangīl bi-d-dini who sup rich in-def-world ‘Who is the richest (person) in the world?’

In Cilicia, comparison is often expressed by the elative pattern of an adjective, which is preceded by the particle . This clearly reflects a : the Turkish equivalent of the adverb issa ‘still, yet’ is daha, which in Turkish is also used as the particle of the comparative.

34For more examples and further details see Procházka (2002a: 200–201) for Cilicia, and Procházka & Batan (2016: 464–465) for the Bedouin-type dialects in the Harran–Urfa region.

102 4 Arabic in Iraq, Syria, and southern Turkey

(33) Cilician Arabic (Procházka 2002a: 202) ṣāyir issa aḥsan become.ptcp more good.ela ‘It became better’. (34) Turkish Daha iyi ol-du. more good become-prf.3sg ‘It became better.’

3.3.2.5 Valency Sometimes a change in verb valency occurs as a consequence of the copying of Turkish models. A case found throughout these dialects is the verb ʕaǧab ‘to like’: usually in Arabic the entity that is liked is the grammatical subject and the person who likes something is the direct object of the verb; but in the Arabic dialects in question, the construction of this verb reflects its Turkish (and English) usage with the person doing the liking the grammatical subject.

(35) a. Cilicia (Procházka 2002a: 200) ʕǧabt bayt-ak like.prf.1sg house-2sg.m b. Damascus (own data) bēt-ak ʕažab-ni house-2sg.m like.prf.3sg.m-1sg ‘I liked your house.’

3.4 Lexicon Apart from the Aramaic loanwords also found in Classical Arabic (see Retsö 2011; van Putten, this volume) – often in the realms of religion and cult – the dialects of this region exhibit a large number of Aramaic lexemes. They are particularly common in Lebanon and western Syria, but also found in Iraq and even in the Bedouin-type dialects (Féghali 1918; Borg 2004; 2008). A large percentage of these words belong to flora and fauna, agriculture, architecture, tools, kitchen utensils, and other material objects:35

35See also Neishtadt (2015: 282). Note that, unless otherwise indicated, lexemes cited in this section are taken from Barthélemy (1935) for Syrian dialects, and Woodhead & Beene (1967) and al-Bakrī (1972) for Iraqi dialects.

103 Stephan Procházka

(36) ṣumd ~ ṣimd ‘plough’ < Syriac ṣāmdē ‘yoke’ qālūz ‘bolt (of a door)’ < Syriac qālūzā nāṭūr ‘guard (of a vineyard etc.)’ < Syriac nāṭūrā šaṭaḥ ‘to spread’ < Syriac šeṭaḥ šōb ‘heat, hot’ < Syriac šawbā

Many nautical terms and words denoting agricultural products and tools were borrowed by Arabic from Greek, often via other languages, especially Aramaic,36 the Lingua Franca, and Turkish:

(37) brāṣa < Greek práson ‘leek’ laḫana < Greek láḫana ‘cabbage’ dərrāʔen < Greek dōrákinon ‘peaches’ ʔabrīm/brīm ‘keel’ < Greek prýmnē ‘stern, poop’ sfīn < Greek sfēn ‘wedge’

Kurdish borrowings are mainly restricted to northern Iraq, where bilingualism is widespread: (38) Mosul pūš ‘chaff’ < Kurdish pûş hēdi hēdi ‘slowly’ < Kurdish hêdî (Jastrow 1979: 68) The intensive cultural and economic contacts between Iraq and Iran led to many Persian loanwords in various domains of the Iraqi dialects.

(39) mēwa ‘fruit’ < Persian mīva ~ mayva baḫat ‘luck’ < Persian baḫt čariḫ ‘wheel’ < Persian čarḫ gulguli ‘pink’ < Persian gol ‘rose’ yawāš ‘slow’ < Persian yavāš puḫta ‘mush’ < Persian poḫte ‘(well) cooked’

Ottoman Turkish contributed a great deal to culinary vocabulary and the ter- minology of clothing and (technical) tools of Syria and Iraq.37 It was even the source of several adverbs and even verbs in the local Arabic varieties (Halasi- Kun 1969; 1973; 1982).

36This is especially true for words related to Christian and ritual, which constitute about twenty per cent of the Greek vocabulary that entered the dialects of Syria. 37The same loanwords are, of course, often found in other regions that were under Ottoman rule, above all in , but also in , and other regions.

104 4 Arabic in Iraq, Syria, and southern Turkey

(40) Syria (Damascus) šāwərma ‘’ < Turkish çevirme ṣāž ‘iron plate for making bread’ < Turkish saç yalanži ‘vine-leaves stuffed with rice’ < Turkish yalancı ‘liar’ (as they pretend to be “real” stuffed with meat) šīš ṭāwūʔ ‘spit-roasted chicken’ < Turkish şiş tavuk kǝzlok ‘glasses’ < Turkish gözlük ʔūḍa ‘room’ < Turkish oda ballaš ‘to begin’ < Turkish başla-mak by metathesis. (41) Iraq (Muslim Baghdadi, cf. Reinkowski 1995) qūzi ‘a dish with roasted mutton’ < Turkish kuzu ‘lamb’ tēl ‘wire’ < Turkish tel yašmāɣ ‘kerchief (for men)’ < Turkish yaşmak ‘veil (for women)’ bōš ‘empty; neutral’, which yielded also the verb bawwaš ‘to put into neutral (gear)’ < Turkish boş ‘empty’ qačaɣ ‘smuggled goods’ < Turkish kaçak

During the last century, the Arabic dialects in Turkey38 have incorporated numerous Turkish words in addition to loanwords from Ottoman times. Among them are terms in education, medicine, sports, media, and technology. Besides these, kinship terms, the vocabulary of everyday life, and structural words like adverbs and discourse markers have infiltrated the dialects from Turkish.

(42) Cilician Arabic qāyin … ‘-in-law’ (< Turkish kayın) ṭōrūn ‘grandchild’ (< Turkish torun) bīle ‘even’ (< Turkish bile) qāršīt ‘opposite from’ (< Turkish karşı)

The cases of semantic extension of an Arabic word result from the wider se- mantic range of its Turkish equivalent which has been transferred into Arabic. Thus, in both Cilician and Harran–Urfa Arabic sāq/ysūq ‘to drive’ also occurs with the meaning of ‘to last’ like the Turkish verb sürmek. In Harran–Urfa b- arð̣ ‘on the place/ground (of)’ has become a preposition/conjunction meaning ‘instead’. This can be seen as an instance of contact-induced grammaticalization (Gardani et al. 2015: 4) under the influence of Turkish yerine ‘instead, in its place’.

38For Cilicia see Procházka (2002a; 2002b: 187–199).

105 Stephan Procházka

(43) Harran–Urfa (own data) al-mille tākl-u b-arð̣ al-laḥam def-people eat.impf.3sg.f-3sg.m in-place def-meat ‘The people eat it instead of meat.’ (44) Harran–Urfa (own data) b-arð̣-in tibči ʔigir āya in-place-link cry.impf.2sg.m read.imp.sg.m verse ‘Instead of crying recite a (Koranic) verse!’

In Iraq, many English words related to Western culture and technology have been, and still are, borrowed into the dialects. The same is true for French in Syria and (particularly) Lebanon (cf. Barbot 1961: 176).

(45) Iraq (words of English origin) kitli < kettle buṭil < bottle glāṣ < glass pančar ‘flat tire’ (< puncture) pāysikil < bicycle māṭōrsikil < motorcycle lōri < lorry igzōz < exhaust (pipe) brēk < brake (46) Syria and Lebanon (words of French origin) gātto ~ gaṭō < gâteau ‘cake’ garsōn < garçon ‘waiter’ sēšwār < séchoir ‘hair drier’ kwaffēr < coiffeur ‘hair-dresser’ ʔaṣanṣēr < ascenseur ‘elevator’ grīb < grippe ‘influenza’

Due to long-term contacts, there are mutual borrowings between the Bedouin and sedentary dialects of the region. This affects not only specific vocabulary of the respective cultures but also basic lexical items. Historically, the sedentary dialects have been much more influenced by the Bedouin-type dialects than vice versa.

106 4 Arabic in Iraq, Syria, and southern Turkey

4 Conclusion

The sociolinguistic history of the regions treated here suggests that the condi- tions for imposition were relatively restrictive and mainly found in contact set- tings with Aramaic, which, over the centuries, has been given up by most of its speakers in favor of Arabic. Thus, it is not surprising that so many features be- yond the lexicon for which contact-induced change can be assumed are related to Aramaic influence. Morphological borrowing is in general relatively rare because it presupposes a high intensity of contact (Gardani et al. 2015: 1). Practically all cases presented in §3.2 corroborate the universal tendencies that: (i) derivational morphology is more prone to borrowing than inflectional morphology; and (ii) nominaliz- ers and diminutives are very frequently represented in instances of borrowed derivational morphology (Gardani et al. 2015: 7; Seifart 2013). On the whole, the Bedouin-type dialects exhibit significantly fewer contact-induced changes than the sedentary dialects. This may be the result of both the Bedouin groups’ no- madic way of life at the fringes of the desert and their tribally organized society, which impedes intense contact with outsiders. The relative infrequency of contact-induced changes in morphology and syn- tax found in the Arabic varieties spoken in Turkey have two main explanations: first, the high degree of complete bilingualism is a very recent phenomenon that only pertains to the last two generations; and second, and probably more impor- tantly, the great structural differences between the two languages, which have impeded both matter and pattern replications. What is still relatively unclear is the degree of historical bilingualism between Arabic on the one hand and Ottoman Turkish, Kurdish, and Persian on the other. Future research would be particularly desirable with regard to Iraq, providing interesting new data on contact-induced changes in multilingual regions like Mosul and Kirkuk, where Arabic, Turkmen, and Kurdish speakers have been in contact for a long time. Also, studies like that of Neishtadt (2015) for should be carried out for Syrian and especially Iraqi dialects with regard to lex- ical borrowings from Aramaic. Another completely under-researched topic is idiomatic constructions, in which the mutual influence of most languages in the region may be assumed.

107 Stephan Procházka

Further reading

There are no studies which treat the subject of contacts between Arabic and the other languages of the whole region covered in this chapter. However:

) Arnold & Behnstedt (1993) is an in-depth study of the mutual contacts between Western Neo-Aramaic and the local Arabic dialects in the Anti-Lebanon Moun- tains of Syria. ) Diem (1979) is a pioneer study of substrate influence in the modern Arabic dialects, though with focus on South Arabia, i.e. outside of the region treated in this chapter. ) Palva (2009) is a very good case study of the diachronic relations between sedentary and Bedouin-type dialects in the Iraqi capital Baghdad. ) Weninger (2011) is a concise overview of contact between different varieties of Aramaic and Arabic.

Acknowledgements

I am grateful to my colleagues Bettina Leitner and Veronika Ritt-Benmimoun for their valuable comments on earlier drafts of this paper. I warmly thank Jérôme Lentin for extensive discussion of the possible origin of the hypocoristic -o suffix (§3.2.4) and his help in finding important sources.

Abbreviations

1, 2, 3 1st, 2nd, 3rd person L1 BCE before Common Era L2 second language CE Common Era link linker comp complementizer m masculine def definite OA Old Arabic f feminine obl oblique ela elative degree pl plural exs existential prf perfect (suffix conjugation) imp imperative sg singular impf imperfect ( conjugation) sup superlative indf indefinite

108 4 Arabic in Iraq, Syria, and southern Turkey

References

Abu-Haidar, Farida. 1991. Christian Arabic of Baghdad. Wiesbaden: Harrassowitz. Abu-Haidar, Farida. 1999. Vocatives as exclamatory nouns in Iraqi Arabic. In Yasir Suleiman (ed.), and linguistics, 144–160. London: Routledge. Aikhenvald, Y. 2004. Evidentiality. Oxford: . al-Bakrī, Ḥāzim. 1972. Dirāsāt fī l-alfāð̣ al-ʕāmmiyya al-mawṣiliyya. Baghdad: Maṭbaʕat Asʕad. Arnold, Werner. 1998. Die arabischen Dialekte Antiochiens. Wiesbaden: Harrassowitz. Arnold, Werner & Peter Behnstedt. 1993. Arabisch-aramäische Sprachbeziehungen im Qalamūn (Syrien). Wiesbaden: Harrassowitz. Barbot, Michel. 1961. Emprunts et phonologie dans les dialectes citadins syro- libanais. Arabica 8(2). 174–188. Barthélemy, Adrien. 1935. Dictionnaire arabe-français, dialectes de Syrie: Alep, Damas, Liban, Jérusalem. Vol. 5. : Paul Geuthner. Volumes 4–5 edited by H. Fleisch. Behnstedt, Peter. 1992. Zur Spaltung von *ā in ō/ē (ǭ/ę̄) im Qalamūn (Antilibanon). Die Welt des 23. 77–100. Behnstedt, Peter. 1994a. Dialektkontakt in Syrien. In Dominique Caubet & Martine Vanhove (eds.), Actes des premières journées internationales de dialecto- logie arabe de Paris, 417–430. Paris: Langues’O. Behnstedt, Peter. 1994b. Der arabische Dialekt von Soukhne (Syrien). Teil 2: Phono- logie, Morphologie, Syntax. Teil 3: Glossar. Wiesbaden: Harrassowitz. Behnstedt, Peter. 1997. Sprachatlas von Syrien. Vol. I: Kartenband. Wiesbaden: Harrassowitz. Berlinches, Carmen. 2016. El dialecto árabe de Damasco (Siria): Estudio gramatical y textos. Zaragoza: Prensas de la Universidad de Zaragoza. Blanc, Haim. 1964. Communal dialects in Baghdad. Cambridge, MA: Harvard University Press. Borg, . 1985. : A historical and comparative investigation into the phonology and morphology of the Arabic spoken by the of Kormakiti village in the Kyrenia district of north-west Cyprus. Stuttgart: Deutsche Morgenländische Gesellschaft. Borg, Alexander. 2004. A comparative glossary of Cypriot Maronite Arabic (Arabic– English): With an introductory essay. Leiden: Brill. Borg, Alexander. 2008. Lexical Aramaisms in nomadic Arabic . In Stephan Procházka & Veronika Ritt-Benmimoun (eds.), Between the Atlantic and Indian Oceans: Studies on contemporary Arabic dialects. Proceedings of the

109 Stephan Procházka

7th AIDA Conference, held in Vienna from 5-9 September 2006, 89–112. Münster: LIT-Verlag. Brockelmann, Carl. 1928. Deminutiv und Augmentativ im Semitischen. Zeitschrift für Semitistik und verwandte Gebiete 6. 109–134. Brustad, Kristen. 2000. The syntax of spoken Arabic: A comparative study of Moroccan, Egyptian, Syrian, and Kuwaiti dialects. Georgetown: Georgetown University Press. Cantineau, Jean. 1960. Etudes de linguistique arabe. Paris: Klincksieck. Coghill, Eleanor. 2012. Parallels in the grammaticalisation of Neo-Aramaic zil- and Arabic raḥ- and a possible contact scenario. In Domenyk Eades (ed.), Gram- maticalization in Semitic, 127–144. Oxford: Oxford University Press. Contini, Ricardo. 1999. Le substrat araméen en néo-arabe libanais: Préliminaires à une enquête systématique. In Marcello Lamberti & Livia Tonelli (eds.), Afro- asiatica Tergestina, 101–122. Padova: Unipress. Corriente, Federico. 1969. Qalqūl en Semitico. Sefarad 29. 3–11. Cowell, Mark W. 1964. A reference grammar of . Washington, DC: Georgetown University Press. Diem, Werner. 1979. Studien zur Frage des Substrats im Arabischen. Der Islam 56. 12–80. Donner, Fred McGraw. 1981. The early Islamic conquests. Princeton: Princeton University Press. Fassberg, Steven Ellis. 2010. The Jewish Neo-Aramaic dialect of Challa. Leiden: Brill. Ferguson, Charles A. 1969. The /g/ in Syrian Arabic: Filling a gap in a phonolog- ical pattern. Word 25. 114–119. Ferguson, Charles A. 1997. Arabic baby talk. In Kirk Belnap & Niloofar Haeri (eds.), Structuralist studies in Arabic linguistics: Charles A. Ferguson’s papers, 1954–1994, 179–187. Leiden: Brill. Féghali, Michel. 1918. Etude sur les emprunts syriaques dans les parlers arabes du Liban. Paris: Champion. Fleisch, Henri. 1974a. Le parler arabe de Kfar-Ṣghâb, Liban. In Henri Fleisch (ed.), Etudes d’arabe dialectal, 221–262. Beirut: Dar el-Machreq. Fleisch, Henri. 1974b. Le changement ā > ō dans le sémitique de l’ouest et en arabe dialectal libanais. In Henri Fleisch (ed.), Etudes d’arabe dialectal, 45–50. Beirut: Dar el-Machreq. Gardani, Francesco, Peter Arkadiev & Nino Amiridze (eds.). 2015. Borrowed mor- phology. Berlin: De Gruyter.

110 4 Arabic in Iraq, Syria, and southern Turkey

Gazsi, Dénes. 2011. Arabic– contact. In Stefan Weninger, Geoffrey Khan, P. Streck & Janet C. E. Watson (eds.), The : An international handbook, 1015–1021. Berlin: De Gruyter Mouton. Gralla, Sabine. 2006. Der arabische Dialekt von Nabk (Syrien). Wiesbaden: Harrassowitz. Grigore, George. 2007. L’arabe parlé à : Monographie d’un parler arabe “périphérique”. Bucharest: Editura Universitãţii Bucureşti. Halasi-Kun, Tibor. 1969. The Ottoman elements in the Syrian dialects. Archivum Ottomanicum 1. 14–91. Halasi-Kun, Tibor. 1973. The Ottoman elements in the Syrian dialects, II. Archivum Ottomanicum 5. 17–95. Halasi-Kun, Tibor. 1982. The Ottoman elements in the Syrian dialects, II (contin- uation). Archivum Ottomanicum 7. 117–267. Holes, Clive. 2002. Non-Arabic Semitic elements in the Arabic dialects of east- ern Arabia. In Otto Jastrow, Werner Arnold & Hartmut Bobzin (eds.), “Sprich doch mit deinen Knechten aramäisch, wir verstehen es!”: 60 Beiträge zur Semi- tistik: Festschrift für Otto Jastrow zum 60. Geburtstag, 269–279. Wiesbaden: Harrassowitz. Holes, Clive. 2007. Colloquial Iraqi Arabic. In . N. Postgate (ed.), Languages of Iraq, ancient and modern, 123–134. Baghdad: British School of in Iraq. Holes, Clive. 2016. Dialect, culture, and society in . Vol. 3: Phonology, morphology, syntax, . Leiden: Brill. Hopkins, Simon. 1995. ṣarār ‘pebbles’: A Canaanite substrate word in . Zeitschrift für Arabische Linguistik 30. 37–49. Hopkins, Simon. 1997. On the construction šmēh l-gabrā ‘the name of the man’ in Aramaic. Journal of Semitic Studies 42. 23–32. Jastrow, Otto. 1978. Die mesopotamisch-arabischen qəltu-Dialekte. Vol. 1: Phono- logie und Morphologie. Wiesbaden: Steiner. Jastrow, Otto. 1979. Zur arabischen Mundart von Mossul. Zeitschrift für Arabische Linguistik 2. 76–99. Jastrow, Otto. 1990. Die arabischen Dialekte der irakischen Juden. In Werner Diem & Abdoldjavad Falaturi (eds.), XXIV. Deutscher Orientalistentag vom 26. bis 30. September 1988 in Köln: Ausgewählte Vorträge, 199–205. Wiesbaden: Steiner. Kahane, Henry Romanos, Renée Kahane & Andreas Tietze. 1958. The Lingua Franca in the Levant: Turkish nautical terms of Italian and Greek origin. Urbana, IL: University of Illinois Press.

111 Stephan Procházka

Khoury, Philipp S. 1987. Syria and the French Mandate: The politics of Arab nation- alism, 1920–1945. London: Tauris. Krebernik, Manfred. 2008. Von Gindibu bis Muḥammad: Stand, Probleme und Aufgaben altorientalistisch-arabistischer Philologie. In Otto Jastrow, Shabo Ta- lay & Herta Hafenrichter (eds.), Studien zur Semitistik und Arabistik: Festschrift für Hartmut Bobzin zum 60. Geburtstag, 247–279. Wiesbaden: Harrassowitz. Lentin, Jérôme. 2018. The Levant. In Clive Holes (ed.), Arabic historical dialecto- logy: Linguistic and sociolinguistic approaches, 170–205. Oxford: Oxford Univer- sity Press. Lucas, Christopher. 2015. Contact-induced language change. In Claire Bowern & Bethwyn Evans (eds.), The Routledge handbook of , 519–536. London: Routledge. Magidow, Alexander. 2013. Towards a sociohistorical reconstruction of pre-Islamic Arabic dialect diversity. Austin: University of . (Doctoral dissertation). Masliyah, Sadok. 1996. Four Turkish suffixes in Iraqi Arabic: -li, -lik, -siz and -çi. Journal of Semitic Studies 41(2). 291–300. Masliyah, Sadok. 1997. The diminutive in spoken Iraqi Arabic. Zeitschrift für Arabische Linguistik 33. 68–88. Muraoka, Takamitsu. 2005. Classical Syriac. 2nd, revised edn. Wiesbaden: Harrassowitz. Neishtadt, Mila. 2015. The lexical component in the Aramaic substrate of Palestinian Arabic. In Butts (ed.), Semitic languages in contact, 280–310. Leiden: Brill. Nöldeke, Theodor. 1904. Neue Beiträge zur semitischen Sprachwissenschaft. Stras- bourg: Trübner. Owens, Jonathan. 2006. A linguistic history of Arabic. Oxford: Oxford University Press. Palva, Heikki. 2009. From qǝltu to gilit: Diachronic notes on linguistic adaptation in Muslim Baghdad Arabic. In Enam Al-Wer & Rudolf De Jong (eds.), Arabic dialectology in honour of Clive Holes on the occasion of his sixtieth birthday, 17– 40. Leiden: Brill. Pat-El, Na’ama. 2017. Digging up archaic features: Neo-Arabic and comparative Semitics in the quest for proto-Arabic. In Al-Jallad (ed.), Arabic in con- text: Celebrating 400 years of Arabic at Leiden University, 441–475. Leiden: Brill. Perry, John . 2007. Persian morphology. In Alan S. Kaye (ed.), Morphologies of Asia and , 975–1019. Winona Lake, IN: Eisenbrauns. Procházka, Stephan. 2002a. Die arabischen Dialekte der Çukurova (Südtürkei). Wiesbaden.

112 4 Arabic in Iraq, Syria, and southern Turkey

Procházka, Stephan. 2002b. Contact phenomena, code-copying, and code- switching in the Arabic dialects of Adana and Mersin (southern Turkey). In Abderrahim Youssi (ed.), Aspects of the dialects of Arabic today: Proceedings of the 4th conference of the International Arabic Dialectology Association (AIDA), , April 1-4, 2000. In honour of Professor Cohen, 133–139. : Amapatril. Procházka, Stephan. 2004. Unmarked feminine nouns in modern Arabic dialects. In Martine Haak, Rudolf de Jong & Kees Versteegh (eds.), Approaches to Arabic dialects: A collection of articles presented to Manfred Woidich on the occasion of his sixtieth birthday, 237–282. Leiden: Brill. Procházka, Stephan. 2013. Traditional boatbuilding: Two texts in the Arabic dialect of the island of Arwād (Syria). In Renaud Kuty, Ulrich Seeger & Shabo Talay (eds.), Nicht nur mit Engelszungen: Beiträge zur semitischen Dialektologie. Festschrift für Werner Arnold zum 60. Geburtstag, 275–288. Wiesbaden: Harrassowitz. Procházka, Stephan. 2018. The northern Fertile Crescent. In Clive Holes (ed.), Arabic historical dialectology: Linguistic and sociolinguistic approaches, 257–292. Oxford: Oxford University Press. Procházka, Stephan & Ismail Batan. 2016. The functions of active participles in Šāwi Bedouin dialects. In & Biţună (eds.), Arabic va- rieties: Far and wide. Proceedings of the 11th International Conference of Aida, Bucharest 2015, 457–466. Bucharest: Editura Universității din București. Procházka-Eisl, Gisela. 2018. A suffix on the move: Forms and functions ofthe Turkish suffix -ci in Arabic dialects. Mediterranean Language Review 25. 21–52. Prunet, Jean-François & Idrissi. 2014. Overlapping morphologies in Arabic hypocoristics. In Sabrina Bendjaballah, Noam Faust, Mohamed Lahrouchi & Nicola Lampitelli (eds.), The form of structure, the structure of form: Essays in honor of Jean Lowenstamm, 177–192. Amsterdam: John Benjamins. Reinkowski, Maurus. 1995. Eine phonologische Analyse türkischen Wortgutes im Bagdadisch-Arabischen mitsamt einer Wortliste. Folia Orientalia 31. 89–115. Retsö, Jan. 2011. Aramaic/Syriac loanwords. In Lutz Edzard & Rudolf de Jong (eds.), Encyclopedia of Arabic language and linguistics, online edn. Leiden: Brill. Rubin, Aaron D. 2005. Studies in Semitic grammaticalization. Winona Lake, IN: Eisenbrauns. Sabuni, Abdulghafur. 1980. Laut- und Formenlehre des arabischen Dialekts von Aleppo. Frankfurt am Main: Peter Lang. Sakel, Jeanette. 2007. Types of loan: Matter and pattern. In Yaron Matras, Georg Bossong & Bernard Comrie (eds.), Grammatical borrowing in cross-linguistic perspective, 15–29. Berlin: Mouton de Gruyter.

113 Stephan Procházka

Seifart, Frank. 2013. AfBo: A world-wide survey of affix borrowing. : Max Planck Institute for Evolutionary Anthropology. http://afbo.info. Shields, . 2004. Mosul : Economy, identity, and annexation. In Reeva S. Simon & Eleanor Harvey Tejirian (eds.), The creation of Iraq: 1914– 1921, 50–60. : Columbia University Press. Singer, -Rudolf. 1984. Grammatik der arabischen Mundart der von . Berlin: De Gruyter. Souag, Lameen. 2017. Clitic doubling and language contact in Arabic. Zeitschrift für Arabische Linguistik 66. 45–70. Talay, Shabo. 1999. Der arabische Dialekt der Khawētna. Vol. 1: Grammatik. Wiesbaden: Harrassowitz. Trudgill, Peter. 1986. Dialects in contact. Oxford: Wiley-Blackwell. Weninger, Stefan. 2011. Aramaic–Arabic language contact. In Stefan Weninger, Geoffrey Khan, Michael P. Streck & Janet C. E. Watson (eds.), The Semitic lan- guages: An international handbook, 747–755. Berlin: De Gruyter Mouton. Wilkins, Charles L. 2010. Forging urban solidarities: Ottoman Aleppo 1640–1700. Leiden: Brill. Wilmsen, David. 2014. Arabic indefinites, and negators: A linguistic history of western dialects. Oxford: Oxford University Press. Woidich, Manfred. 2006. Das Kairenisch-Arabische: Eine Grammatik. Wiesbaden: Harrassowitz. Woodhead, R & Wayne Beene. 1967. A of Iraqi Arabic: Arabic– English. Washington, DC: Georgetown University Press.

114